This action might not be possible to undo. Are you sure you want to continue?
Authors: Robert A. Halvorson, SE, PE, FIStructE, FASCE, Principal Halvorson and Partners, 600 West Chicago Avenue, Suite 650, Chicago, Illinois 60654 312-274-2402 firstname.lastname@example.org Carrie Warner, SE, PE, LEED AP, Associate Principal Halvorson and Partners, 600 West Chicago Avenue, Suite 650, Chicago, Illinois 60654 312-274-2413 email@example.com Alex Lang, Staff Engineer Halvorson and Partners, 600 West Chicago Avenue, Suite 650, Chicago, Illinois 60654 312-274-2400 firstname.lastname@example.org
The 600 m (1968 ft) tall Russia Tower has an extremely efficient and intuitively clear structural concept. But its unique form raised a number of key structural challenges to be addressed – all related to stability. The following paper outlines Halvorson and Partners’ approach to these: applying stability concepts from 2005 AISC Specification for Structural Steel Buildings (AISC, 2005), including the newly introduced Direct Analysis Method (DM), as well as material secondary affects for concrete from ACI to assess, analyze, and design for stability in this complex, composite structure – the Russia Tower.
Tower Structure Overview
The Russia Tower will be Europe’s tallest building and one of the most distinctive high rises in the world. Its Y-shaped plan extrudes to 600 m, tapering in elevation, to create a visually slender form of three 23 m (75 ft) wide wings extending from a central spine. The expressed “braced spine” structure developed by Halvorson and Partners, in collaboration with architect Foster + Partners, is an an extremely efficient concept for super-tall structures. A complete description of the tower’s structural concept, components and collaborative design process can be found ASCE 2009 paper by Robert Halvorson, entitled “Russia Tower: Design Challenges”.
2008) The central spine of the structure consists of reinforced concrete walls hidden in core elements – which are linked together across the open space in the central spine by two-story steel chevron bracing. On each wing face. with no concerns for uplift at the base. 2 .FIGURE 1. only span like a continuous beam between the points where the fan columns brace the spine. in effect.25 m2 (10 to 67 ft2). Lateral loads are resisted as axial forces in the sloped columns and the ‘spine’ must. 2007) Structural Diagram ( Halvorson and Partners. These reinforced concrete columns.0 to 6. where they then “bounce off” the spine to create the reverse fan columns (see figure 1). which vary in cross-sectional area from 1. substantial horizontal reinforced concrete beams are provided to resist the thrusts of the fan columns. Steel floor framing interfaces with the steel bracing directly. creating ‘closed sections’ in each wing for torsional stiffness. Gravity loads are also resisted as axial forces in the sloped columns and walls. Where the fan columns intersect the spine. The structural system is extremely efficient. and connects light steel erection columns within the concrete walls and sloped columns. gravity stresses exceed any wind tensions in all but a few stories of select columns. And lateral displacements and accelerations are well below standard limits. Russia Tower Rendering (Foster + Partners. Four-story steel chevron bracing at the wing tips serve to link the rigid faces formed by the fan and reverse fan columns. brace the central spine for lateral loads and carry gravity loads. seven fan columns radiate from a base abutment toward the central spine. Given this and its low aspect ratio.
5) Flexible floor diaphragms. of this “Y” shaped form. Halvorson also retained Dr. The resulting approach incorporated three separate. the bracing with nodes at variable intervals. ACI’s criteria for concrete structures have not changed in some time and they similarly recommend general requirements for safety and stability. such as out-of-plumb (yet still within tolerance) construction. Second. but it was necessary to evaluate global buckling. Chapter C “Frames and Other Structures” provides clearer requirements for general stability. and. A ‘rigorous’ second order analysis was developed to include: 1) Large P-Delta (global) and small p-delta (local member) effects. giving the engineer more guidance and at the same time allowing more flexibility in addressing this issue. 4) Material non-linearity. sections 10. and. First. Jerry Hajjar of the University of Illinois to assist in developing an analysis approach appropriate to the unique characteristics of the Russia Tower. Stability of the highly stressed concrete fan columns resulting from uncertainty on how to interpret their unbraced lengths according to conventional design methods. 2) Flexure. 3) Geometric imperfections. but closely related. Axial and Shear Deformation. Halvorson called upon guidance from both AISC and ACI. particularly in torsion considering the “Y” shape of the tower. “notional” loads are defined to address geometric imperfection. Global stability. previously. the new Direct Analysis Method in Appendix 7 presents a clear and encompassing methodology that will likely soon become the standard for design. this was handled less transparently. Some more specific guidance is offered in section ACI 318-05. especially for torsion. 2. Global Stability Analysis Stability Analysis Concepts: Russia Tower’s form provides a very stable base for resisting lateral loads. analyses and design checks to address each of the three issues.11. Third.10 and 10. and so forth). This analysis needed to accurately reflect the flexibilities in the system and consider all secondary effects. The 2005 AISC specification has greatly expanded and refined its treatment of stability in steel structures. and perhaps most notably. resulting in reduced stiffness of members due to residual stresses in steel and creep and shrinkage in concrete. This method allows all columns to simply be designed using an unbraced length equal to one story height. Given its composite configuration.Stability Challenges With the unique configuration of this structural system (cruciform plan. for considering secondary material affects such as creep and shrinkage by modifying member and material properties. and 3. Design of the floor diaphragms to insure adequate strength and stiffness to maintain the plan configuration of the tower. there were a number of key structural challenges to be addressed– all related to stability: 1. 3 . to laterally brace the columns.
each time increasing the factored loads until an eigenvalue of 1.2% of the gravity loads applied to each column at every floor acting horizontally. Buckling cases with wind were conservatively considered. Elements can be assessed for extent of cracking and its affect on behavior to arrive at approximate factors. both ACI and AISC offer guidance for this. For the Russia Tower. However. notional loads. 0. The value for the notional load suggested in AISC. The full tower model analyzed in SAP2000 accurately represented all primary framing and wall elements for the lateral system. these loads can be orientated to most de-stabilize the structure. notional loads causing the tower to twist were most critical.0 was obtained – representing incipient buckling. those inducing sidesway were also separately considered. Horizontal notional loads can be applied to the structure to address Item 3. This non-linear (non additive) method was used in conjunction with a linear buckling analysis as later described. with openings conservatively represented such that flexibility of the floor diaphragms was modeled. but this was not necessary during assessment of global buckling. Gravity loads. although there are still some areas of ambiguity. A second order analysis was first run under factored loads and its results were used as input for the linear buckling analysis.Items 1 and 2 can be handled by most of the high-end analysis programs available today. This assumption is conservative for tall structures since there is an upper limit on the horizontal offset of columns per AISC. represents a condition wherein all columns and walls are constructed out of plumb (or in this case off of the specified slope!) by height/500 to match typical AISC tolerance limits. the p-Delta plus large displacement method offered in SAP2000 was adopted. This is an area where ‘engineering judgment’ comes into play.0 was reached. but wind loads were not taken simultaneously with notional loads. and not assuming typical ‘rigid diaphragms’. Item 4 can most easily be considered by applying modifiers to members and materials. Wind loads are not usually considered in buckling analysis because this load is transient and would not induce material non-linear effects. and wind loads were all considered in factored load combinations per ASCE 7-02. since the gravity and wind loads are both carried as axial loads in the same members. a non-linear buckling analysis was carried out under factored loads. Stability Analysis Methodology: In carrying out the study to assess the tower stability. including perimeter beams. Small p-delta could be accounted by dividing members into two pieces. In line with new AISC recommendations. Analysis was done in an iterative process. sequentially increasing these loads until a lowest eigenvalue of 1. Item 5 can be captured by accurately modeling floors. such as creep. For the buckling analysis and investigation. Three directions of notional loads were considered: two in lateral directions and the third in a torsional manner. The ratio of the final loads to the required factored loads represents the true non-linear buckling Eigen value ratio. This two-step process was carried out multiple times. Floors were modeled with slab elements. this was conservatively done. 4 . however.
For the slab itself. ACI recommends a 1 / (1 + bd) modifier.0 would not be reached until the factored load combination was increased by a factor 2.625 x E concrete). The columns’ I modifier was established to insure EI = 0. the lowest.7 times the required factored design loads. modifiers were applied to material and member properties. An eigenvalue of 1. non-linear buckling occurred) when this load combination was increased to approximately 1.0 was achieved (i. in line with AISC Appendix 7. Based on this analysis. Similarly. considering gravity loads acting alone. buckling mode was found for the case considering gravity loads plus 100% of torsional wind loads acting together with a portion of lateral wind loads. and = 0.50 perpendicular to the wing. Where steel framing and diaphragm bracing provided additional diaphragm stiffness. which Halvorson elected to apply this ratio to Econcrete (0. an eigenvalue of 1. or roughly a factor of 3 times actual loads. For concrete buckling assessment. or roughly 2. which were provided with sufficient reinforcing. the Modulus of Elasticity was taken as 0.1 times actual loads (see Figure 2). =0.50 modifier on Ig for link beams. floor framing and concrete slabs. further in line with ACI recommendations for buckling. For steel. this was also included in the full model buckling analysis. Therefore.1 x service loads.20 in both directions within the centre where the wings meet). such as a 0. only the concrete above the flutes was considered for in-plane stiffness.e. Torsional buckling of tower at ratio of 1. and therefore governing.. to insure the gravity arching action was carried in the perimeter beams and not the diaphragm.4. Other concrete members modified based on the effects of cracking. The floor diaphragm stiffness needed to represent contributions from steel diaphragm bracing. FIGURE 2.20 E Ig.10 along wing.80 x E steel. with reductions applied for cracking (In-plane stiffness modifier = 0.To capture material non-linearity. it was concluded that the tower was stable in a global sense. 5 .7 x factored loads (including wind). or 2.
for reasonably small 6 . Four-story portion of tower showing varied modules of bracing AISC’s Direct Analysis Method for steel structures offered a method for handling this issue. Every effort was made to minimize their size while insuring the strength and stability of these highly loaded members. the iterative approximated p-Delta method in ETABs was sufficient for capturing deformations and secondary forces. with the following exceptions outlined by previous numbered items: 1) For capturing Large P-Delta (global) effects. four-story bracing at the wing tips. By carrying out an analysis that considers all secondary effects. The final designs of the columns were taken as the more conservative of the two. see Figure 3). Direct Analysis Method: The concepts and modeling assumptions applied for the Direct Analysis Method are nearly identical to those outlined previously for the Global Buckling Analysis. the Direct Analysis Method allows all columns be designed with k=1. Since the fan columns were apparently braced at varied modules (core walls continuous at all levels. gravity support for the tower. Their visual expression is a (if not ‘the’) defining architectural feature. two-story bracing at the spine between the cores. and sloping columns triangulating over varied story heights. and served as a model for the method developed for the Russia Tower’s composite structure.Column Stability and Design The sloping fan columns are the key to Russia Tower. it was challenging to determine the unbraced column lengths for use in a conventional design approach. meaning considering a one-story column height. along with the core walls. the strength analysis and design was actually carried out in two versions: a Direct Analysis Method and a Conventional Method. subject to the guidelines given in AISC’s Appendix 7. the team concluded that for the design of the Russia Tower. FIGURE 3. They are the primary lateral bracing and. In the final design of the tower.
But. 2) Flexure. additional moments were included in the column design for small p-delta effects in lieu of dividing all columns into two finite elements in the analysis. Using this method.3% gravity loads were instead used. The resulting minimum k values became k=4 for fan columns perpendicular to the wing and k=2 for all other members in all other directions. Except concrete Fan and Reverse Fan Column Moment of Inertias were assumed I effective = 0.<1. in versions with and without springs at the base to bracket potential differential settlement design effects. 5) Flexible floor diaphragm assumptions aligned with those earlier described. 4) Material and members modifiers were identical to those previously outlined for Global Stability.5 (usually <1. Also. and in line with ACI standards. the columns were conservatively considered with k=2 for all cases. since columns were highly stressed and in lieu of calculating tb. and a second sequential model was analyzed in SAP2000 considering non-linear creep and shrinkage effects on concrete. This conventional model was analyzed for gravity and wind loads. a ‘conventional’ design was also carried out in accordance with ACI standards. As a final conservatism. Member modifiers remained similar to the Direct Analysis Method as these were in line with typical ACI standards. Additional moments were again included in column design for local member p-delta effects. However. 7 . making it excellent for design.70 I gross. since D2nd/D 1st.1). Axial and Shear Deformation could be accounted for by ETABS analysis. given the presence of openings and potential for future tenant openings already established by the design team.deformation this approximated iterative method was warranted sufficiently accurate and this method also conveniently allows load cases to be added. were still not required to be included in the factored combination with wind loads for design. a comparison of 2nd and 1st order analyses found D2nd/D 1st. where E varies with time. Small p-delta forces could also be included.5. Independent bucking studies were carried out on four-story modules of the building (established by the height of the four-story tip-bracing module) to establish reasonable estimates for unbraced column lengths. although members may require division into two pieces or including a moment multiplier for these. all columns connected at a floor could be designed for a single story height (k=1). This design and analysis was still done in ETABS.<1. notional loads were no longer included. capturing secondary effects due to p-Delta (Global) and deformations. 3) Notional loads of 0. A version was analyzed in ETABS to consider the potentially different distribution of gravity loads due to construction sequence. but E was no longer reduced. This was an increase from buckling checks. Conventional Method: For comparison. However.
Floor Diaphragm Design In the previous sections on global stability and column design.The final column designs for the Russia Tower represent the most conservative design for each member from all model versions given. carried through axial force in spandrel beams 2) Horizontal wind loads. both for stabilizing the columns to one another and transferring lateral forces to the core walls and fan columns. 8 . the need to provide adequate stiffness and strength in the floor diaphragms to maintain the tower plan form and to laterally brace the columns was raised. This final section explains how these slabs were analyzed and designed for the required in-plane diaphragm forces. The remaining issue was to provide adequate strength in the floor diaphragms. transferring forces into the primary lateral framing members: walls and sloped columns 3) Horizontal forces to brace columns to one another. Diaphragm stresses due to horizontal loads Three primary horizontal forces are acting within the typical floor diaphragm: 1) Horizontal component to stabilize the sloping columns. the most conservative case would be all columns being out of alignment in parallel and all kinking at the same floor level. Around major openings. The earlier analyses evaluated the tower structure considering realistic stiffness of the floor diaphragms. with successful results. FIGURE 4. steel floor framing and added bracing members work together as horizontal trusses transferring in-plane diaphragm forces.
By integrating ACI requirements as explained. analysis and design for the stability of Russia Tower. If a column slopes at height/800 above a level.0025). the forces created by wind loads are of significantly lower magnitudes than the forces required to brace the columns. Halvorson and Partners arrived a reasonably conservative approach to feel confident in our assessment.. etc. pulling a cantilevered wing away from the tower. the third load case was considered for the diaphragm design. Therefore. Steel framing and diaphragm bracing and their connections were designed to resist the axial loads due to these in-plane forces. including the introduction of the Direct Analysis Method. For Russia Tower.25% of the axial load in the column (e. 1/800+1/800=0.g. fundamental-based methodology for addressing such issues. Summary Stability was the key design consideration for the unique configuration of the Russia Tower. “kinks” at that level. 9 . The diaphragm forces in the steel framing and in the concrete slabs were bracketed for two extremes – one with the slab stiffness “turned off” and only the steel diaphragm bracing considered and another with the steel diaphragm bracing omitted and only the concrete floor slab considered. The recently updated stability sections of the AISC specification. a tolerance limit of height/800 was set for the columns. and then slopes at height/800 below in the opposite direction. Of the other two types of diaphragm forces.The first type of diaphragm force created by stabilizing the sloped columns is resisted by a load path provided in the perimeter spandrel beams (for this analysis. Reasonable amounts of reinforcing were sufficient in the concrete slabs to handle the inplane diaphragm forces. the required horizontal force acting at that level to brace the column is equivalent to 0.0025 times the total axial load in the columns of one wing was applied horizontally to one floor level of the tower at a time in various orientations . the concrete slab properties were significantly reduced so as to eliminate load shedding from the spandrel beams into the concrete slab).pushing a cantilevered wing sideways. This bracing force of 0. The analysis revealed that only a portion of the bracing forces had to be resisted by the floor diaphragm at the level where the loads were applied – a significant portion of the bracing forces transferred vertically through the stiff columns and walls to levels above and below. were a timely development for our effort – as these offer engineers a practical.
. ACI 318-05 (2005). Chicago. Inc.References [American Institute of Steel Construction. W. Wisconsin.Proceedings of the Structural Stability Research Council Annual Technical Session and Meeting. American Concrete Institute. (AISC) 2005] “Code of Standard Practice for Steel Buildings and Bridges”. “Building Code Requirements for Structural Concrete and Commentary”. Inc. Pennsylvania. MI. J. (AISC) 2005]. Chicago. 1993. April 6-7. Hajjar. Milwaukee. IL. Bethlehem. [Abdelrazaq. [American Institute of Steel Construction. 10 .. IL. 1993] "Column Buckling Considerations in High-Rise Buildings with Mega-Bracing. 155-169. and Sinn. W. Baker." Is Your Structure Suitably Braced?. A. SSRC. F.. Detroit. pp. “Load and Resistance Factor Design Specification for Structural Steel Buildings”.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.